| { | |
| "best_metric": 0.9357798165137615, | |
| "best_model_checkpoint": "./fine-tune/roberta-base/sst2/checkpoint-4210", | |
| "epoch": 4.0, | |
| "global_step": 16840, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 3.9588281868566905e-06, | |
| "loss": 0.5645, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 7.917656373713381e-06, | |
| "loss": 0.3021, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 1.1876484560570072e-05, | |
| "loss": 0.2658, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 1.5835312747426762e-05, | |
| "loss": 0.2706, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 1.9794140934283453e-05, | |
| "loss": 0.2508, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.9760448779501697e-05, | |
| "loss": 0.2557, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 1.95077576186385e-05, | |
| "loss": 0.2578, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 1.925506645777531e-05, | |
| "loss": 0.2287, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9357798165137615, | |
| "eval_loss": 0.23143097758293152, | |
| "eval_runtime": 1.9026, | |
| "eval_samples_per_second": 458.311, | |
| "eval_steps_per_second": 57.289, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 1.9002375296912114e-05, | |
| "loss": 0.2229, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 1.874968413604892e-05, | |
| "loss": 0.2084, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 1.849699297518573e-05, | |
| "loss": 0.2047, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 1.8244301814322537e-05, | |
| "loss": 0.189, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 1.7991610653459345e-05, | |
| "loss": 0.1914, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 1.773891949259615e-05, | |
| "loss": 0.1874, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "learning_rate": 1.7486228331732958e-05, | |
| "loss": 0.1942, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "learning_rate": 1.7233537170869766e-05, | |
| "loss": 0.1959, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.926605504587156, | |
| "eval_loss": 0.30272433161735535, | |
| "eval_runtime": 1.8407, | |
| "eval_samples_per_second": 473.726, | |
| "eval_steps_per_second": 59.216, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 1.698084601000657e-05, | |
| "loss": 0.1783, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 1.672815484914338e-05, | |
| "loss": 0.1453, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 1.6475463688280183e-05, | |
| "loss": 0.1481, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 1.622277252741699e-05, | |
| "loss": 0.1564, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 1.59700813665538e-05, | |
| "loss": 0.1457, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 1.5717390205690607e-05, | |
| "loss": 0.15, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 1.5464699044827415e-05, | |
| "loss": 0.157, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 1.521200788396422e-05, | |
| "loss": 0.1531, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "learning_rate": 1.4959316723101027e-05, | |
| "loss": 0.1635, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.930045871559633, | |
| "eval_loss": 0.30217334628105164, | |
| "eval_runtime": 1.7605, | |
| "eval_samples_per_second": 495.309, | |
| "eval_steps_per_second": 61.914, | |
| "step": 12630 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "learning_rate": 1.4706625562237835e-05, | |
| "loss": 0.1167, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "learning_rate": 1.4453934401374641e-05, | |
| "loss": 0.1147, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "learning_rate": 1.420124324051145e-05, | |
| "loss": 0.1123, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "learning_rate": 1.3948552079648254e-05, | |
| "loss": 0.1173, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "learning_rate": 1.3695860918785062e-05, | |
| "loss": 0.1105, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "learning_rate": 1.344316975792187e-05, | |
| "loss": 0.1148, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "learning_rate": 1.3190478597058676e-05, | |
| "loss": 0.1335, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "learning_rate": 1.2937787436195484e-05, | |
| "loss": 0.1148, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9288990825688074, | |
| "eval_loss": 0.31621453166007996, | |
| "eval_runtime": 1.6764, | |
| "eval_samples_per_second": 520.163, | |
| "eval_steps_per_second": 65.02, | |
| "step": 16840 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 16840, | |
| "total_flos": 1.772026646744064e+16, | |
| "train_loss": 0.19015042130567683, | |
| "train_runtime": 1734.8381, | |
| "train_samples_per_second": 388.215, | |
| "train_steps_per_second": 24.267 | |
| } | |
| ], | |
| "max_steps": 42100, | |
| "num_train_epochs": 10, | |
| "total_flos": 1.772026646744064e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |